ZigBee network is not online again and again and again

For about 10 days now, I've been getting chronic "ZigBee network is not online" messages in Hubitat, which I usually don't notice until some lights don't start triggering on motion.

This didn't start happening until after we had a power outage here, so I suspect something went wonky then, but other than that, I haven't been messing with Hubitat much at all. In fact, I've been trying to pare down apps and third-party drivers a bit in a vain effort at troubleshooting.

I have mostly Z-wave stuff, so only a few Zigbee motion sensors, a couple of Sengled lights -- 8 Zigbee devices total (aside from about 10-12 Philips Hue bulbs).

I've reset the Zigbee network (full power cycle) DOZENS of times now, and it often goes back offline within an hour.

Last time I rebooted, I opened the "Zigbee Logging" page, and it managed to log only this before crapping out again. Means nothing to me:

Entry 22020-08-03 16:54:29.414 profileId:0x0, clusterId:0x13, sourceEndpoint:0, destinationEndpoint:0 , groupId:0, lastHopLqi:255, lastHopRssi:-61

Entry 12020-08-03 16:54:21.099 profileId:0x0, clusterId:0x13, sourceEndpoint:0, destinationEndpoint:0 , groupId:0, lastHopLqi:255, lastHopRssi:-56

Is there something else I can try? It's really really annoying to have a "smart" house that keeps ignoring me, not to mention the very very low WAF.



Screen shots are always clearer and preferred if possible. Did you see anything in the logs?

Do you mean pull the plug? Or a shut down and reboot? Pulling the plug causes database corruption so each time you do it it gets worse. Luckily there is a fix, you take a backup from the hub and do as soft reset, this wipes the hub apart from the radios. Then when you shut it down and repower it you import your nice and clean back up and your away again.

This sits right with this. If possible put the hub on backup and shut it down correctly on a power outage.


Invest in a UPS. There are several threads on this subject on the forum to give you some ideas.

They did something to zigbee that broke it and will not confess. I have had the same exact thing start happening to my hub and the only thing that "fixed" it was to move all my lightify and tradfri bulbs off of hubitat and back on to smartthings. With all those bulbs gone (they used to work perfectly fine before), my hub has been up a week without a reboot so far.

When? Which version did the change happen in do you think? I'm asking since I've not seen a change in behavior and I've been logging all my Zigbee traffic with Wireshark the last few weeks. If there is a change it would be in how well the controller copes with some type of traffic before croaking, with what I've done to my hubs lately I would think I should have seen it. Don't take this the wrong way, I'm not disbelieving you, but I do know that for 2.2.2 they said straight out that there has been 0 changes to the Zigbee stack. That doesn't mean something else couldn't have changed that has as a side-effect that things crash. There's still things which needs improvements in the Zigbee implementation of HE, I'm personally painfully aware of that, one thing will be fixed for 2.2.3, but there is plenty more to optimize and fix. If this type of failing truly is one of those issues it would be good to understand why.
This being summer I do suspect heat could be another change compared to before as well, but that seems to have been ruled out in similar threads.


Several months ago. I can't pinpoint the exact date for you, sorry. But my system had been working fine until then. I hadn't changed anything except updating the hub and maybe adding some routines, and all of a sudden it started disabling the radio on a daily basis. Not sure why as there are no hub OS logs to look at. But as I have mentioned, moving the lightify and tradfri bulbs off "resolved" the issue. I really wish I knew why as well.

Interesting. I would have never thought wireshark could sniff zigbee as well. What are you using for this? an xbee?

1 Like

Did you ever contact support around that time? They have access to some additional logs from what I understand.

I have a CC2530 USB stick with a +20dBm antenna, it reliably captures all traffic in my whole home. I actually have several in case I need to sniff traffic at multiple channels at the same time. All in all I think you can get the stick, CC Debugger for flashing the stick and the adapter to connect to the debugger for around 15USD on Ali. I bought it all locally here in China though.

This bugs me that they said this because of the amount of people posting daily here with zigbee issues. But I really suspect they changed something else that is effecting zigbee. Like for example, killing a process that is eating up too much resources, and that happens to be zigbee.

Yes, but I was busy and did not get back with them fast enough so they closed my case. The only times I had to troubleshoot were the evenings so I ended up just doing it myself by removing the devices that kept dropping off.

Regarding your heat mention. That was also an issue that made things worse. But it was a side effect of the hub having issues. My hub would go through these cycles on a daily basis:
-Things would get noticeably slow to respond, some not at all.
-The hub would disable zigbee.
-The hub will lock up completely.

I moved the hub to a cooler location which delayed the lockups but they still happened every couple of days instead of multiple times a day.
I performed a hub soft reset which completely resolved the slowdown and lockup issues. (it probably had corruption caused by all the lockups)
All was left was the zigbee radio getting disabled daily. Moving those aforementioned bulbs back to smartthings resolved that issue.
At this time my hub appears to be happy. I'm just wishing I would have gone fully z-wave for hubitat :wink:

1 Like

With the upcoming Radio-backup functionality it would be very interesting to see if a hub with these types of issues being restored to a new hub would have the same issues. Bulbs can be very chatty and traffic-wise very misbehaving as well as cause packet storms due to not answering ACKs according to specs etc. If that makes the radio to be disabled there is something seriously wrong with how things are implemented on HE, but it also is something that should not be left without trying to minimize anyway. I'd be very interested in seeing the Zigbee traffic from a mesh that causes the hub to do this.

I'm all in with Zigbee and WiFi and will probably stay that way. I did just see a Chinese company making some nice-looking Z-wave stuff that runs on the frequency for China, so might start down that road on a trial basis.

Really? Why have there not been hundreds of angry threads that Zigbee is broken? And you think they purposely make it crash? Why would they want that to happen, to anger customers? That makes no sense.

Me neither. I have well over 55 Zigbee devices that vary from Iris to Aqura/Xiaomi to Samsung and more. And none of it seems to give trouble, in fact it has seemed more stable in the last few updates than ever.

Maybe I am not seeing it, but Zigbee seems to draw few complaints. At least I haven't seen widespread outcry that Zigbee is broken.


There were quite a few people, myself and @markus included, who had zigbee issues recently. It occurred just after a hub platform update.

Nobody said anything about purposely marking it crash. The context is a change to zigbee (although Mike Maxwell ruled that out), that broke something, not purposely breaking something :grinning:

1 Like

Well, I've done two soft resets -- the first, restoring a backup from 8/2/20. After restore, the Zigbee network was offline again within the hour.

This morning, I did another soft reset -- this time, restoring a backup from 7/25/20 (still after the initial problems, unfortunately, but the oldest usable backup I had).

This time, the Zigbee network was offline by the first page reload after the restore. sigh

Any other suggestions? Could removing/reinstalling my few Zigbee units help, or is this just a bad hardware issue?

I'll be going out of town soon, and really can't afford to have my network be unreliable while I'm gone. There are cats to feed!

What is the ambient room temperature that the Hubitat Hub is located in? Any chance of it getting a little too warm? I wonder this as I have seen other reports of Zigbee going offline when the hub was in a somewhat warmer than ideal environment.

As a test, would it be possible for you to aim a cooling fan at the bottom of the hub? It would be interesting, as a test, to see if this makes any difference.

Another thing to try is replacing the power supply with a good quality 5v, 1-2amp power supply. Some have reported strange hub issues that were resolved after replacing the power supply. (I personally have been using POE splitters to power my hubs.)


Last night was my first ever Zigbee Offline issue. No amount of rebooting fixed the issue. However, I was running HubConnect between Smartthings V2. I selected the "suspend communication" option and rebooted. Problem solved... sort of. Now my Harmony remote doesn't function with Hubitat, but that was expected. Still troubleshooting.

I've also seen issues after a recent update. Two of my zigbee switches are offline, and I can't get them to come back. Two of the same brand, the other ones still seem to be functioning.

I also had issues with the update breaking my LIFX wifi bulbs, but I reset those and they're ok now.

Also had issues with my Zwave lock, and that is intermittently not responding now too. I'm trying a zwave repair on that, but we'll see.

I'm not clear on which specific update it was, as I might have skipped a few updates. Definitely something in the 2,2.x.x+ range

Same issue here. Overall, the whole platform and increasingly becoming less and less reliable and getting slower. :frowning:

I'm not wishing to be controversial but both my hubs are now running quicker as I can see with my Hub Watchdog stats.
I know if you are having issues thus is not what you want to hear and it can be very frustrating.
I was getting mega issues in the early days which was down to an app and I was very close to forgetting HE and going back to ST. I'm glad I didn't.